Modern technological production systems must achieve high product quality, increased production speed, and reduced waste simultaneously. Traditional manual tuning approaches often fail to balance these conflicting objectives due to the nonlinear and dynamic nature of industrial processes. This study aims to develop an AI-based framework that integrates Machine Learning Regression with Multi-Objective Optimization to identify optimal process parameters in manufacturing. Machine learning models are trained to predict product quality metrics based on process variables, while optimization algorithms determine the best trade-off between quality, throughput, and waste reduction. Experimental evaluation demonstrates that the proposed hybrid system improves prediction accuracy, increases production speed, and significantly reduces defects. This work highlights the potential of AI-based optimization to enhance stability, sustainability, and efficiency in technological processes.